The Future Sustainability of Jute and Tea Production in Bangladesh Under Changing Climate Conditions Using Machine Learning Models

Climate change is expected to impact crop yields globally, especially for economically important crops like tea and jute. This study combines historical yield and climate data with future temperature projections (SSP5-8.5) to forecast yield changes for 36 tea-producing and 20 jute-producing countries by 2050. Using machine learning, Random Forest was the best model to predict tea yields ($\mathbf{R}^{2} \boldsymbol{=} \mathbf{0. 9 4 3 4}$), and $\mathbf{K}$-Nearest Neighbors performed best for jute ($\mathbf{R}^{\mathbf{2}} \boldsymbol{=} \mathbf{0. 9 6 9 4}$). Projections show a decline of around 13.35% in tea yields, while jute yields remain mostly stable due to their broader temperature tolerance in Bangladesh. However, future risks from rainfall changes and water availability may still affect jute. These findings underline the need for crop-specific climate adaptation strategies to support agricultural resilience in a warming world.